Design of embedded TCAM based longest prefix match search engine
Microprocessors & Microsystems
| Hi-index | 0.00 |
TCAM is a popular device for fast routing lookups. TCAM can achieve 0(1) lookup complexity with parallel technique. However, TCAM has three explicit disadvantages: high cost, high power consumption and complex update. For load balancing and policy routing, routers have to hold considerable multi-nexthop routes with multi paths for a destination network. This paper proposes a fast TCAM-based lookup scheme that can support multi-nexthop routing lookups. Using two level index tables, the scheme can store and retrieve multi-nexthop routes quickly. To improve update performance of TCAM, we also propose a fast update algorithm--N-subspace algorithm that can approximately reach 0(1) update complexity for TCAM routing tables. To decrease cost and power consumption of TCAM, we also apply an effective routing compaction method that is based on the trie structure and can reduce 20% routes for current Internet routing tables. Also, the scheme can easily scale to IPv6.